This application addresses broad Challenge Area 05, Comparative Effectiveness Research, specific Challenge Topic AG-102, Prevention and Risk Factor Reduction Strategies for Disabilities. The NIA contact is Ms. Georgeanne Patmios, 301-496-3138, patmiosg@nia.nih.gov. Challenge Topic 05-AG-102 calls for randomized controlled trials (RCTs) to evaluate the comparative effectiveness of competing interventions or modes of intervention delivery. A prime target for 05-AG-102 is the prevention of cognitive disability that results in health outcomes including improved quality of life, decreased mortality, morbidity, and disease progression, reduced medical care costs, and improvements in selected social and behavioral dimensions. The largest and most rigorous RCT ever conducted involving long-term follow-up was the Advanced Cognitive Training for Independent and Vital Elderly (ACTIVE) study. Although all three ACTIVE cognitive training interventions (memory, reasoning, and speed of processing) were effective at improving their targeted abilities, the speed of processing group had the largest gains, with 87% of participants showing reliable improvement. We have also shown clinically significant effects of ACTIVE's speed of processing intervention (but no effects for the memory or reasoning interventions) on a variety of health outcomes, including: (1) a 3% reduction (p = .012) in predicted medical expenses;(2) a 38% reduction in the risk of global decline in health-related quality of life (HRQoL;p = .004);(3) a 30% reduction in the risk of worsening depressive symptoms (p = .012);(4) a 38% reduction in the risk of the onset of suspected clinical depression (p <.01);(5) improvements in self-rated health that translate to a 0.8% absolute reduction in the five-year mortality rate and a 10% relative mortality reduction (p <.05);and, (6) a 64% greater likelihood (p <.05) of meaningful improvements in internal locus of control. Despite the magnitude, diversity, and endurance of these effects of the speed of processing intervention, further research is needed before widespread dissemination is warranted for three reasons. First, ACTIVE relied on a no-contact control group rather than an attention control group, raising the potential for placebo effects. Second, although booster training was randomly offered to 60% of ACTIVE participants, it was offered conditional on completing >8 of the 10 baseline training sessions, confounding booster effects with adherence effects. Third, a new, value-added version of the speed of processing software is now available that can be used on almost any home computer, and could thus dramatically reduce delivery costs and facilitate individual dosing and ongoing booster maintenance, but there is no published evidence that the value-added version is as effective as the original. Therefore, our specific aims are to overcome these limitations using an RCT with one-year follow- up that can be fully completed within the NIH Challenge Grant two-year period. We will randomize 900 participants aged 50 years old or older to three groups. Group G1 (N=400) will receive the value-added speed of processing intervention in 10 onsite sessions as in ACTIVE, with further randomization to one half (G1a) not receiving booster sessions and one half (G1b) receiving onsite booster sessions at 11-months. Group G2 (N=250) will be the attention control group and will receive 10 onsite sessions using a computerized cross-word puzzle program. Group G3 (N=250) will be shown how to operate the value- added speed of processing software on site, and will then be sent home to use it as often as they wish on their own personal computer. Our primary outcome measure is speed of processing, and we will use several reliable and valid instruments to provide a multidimensional assessment, including the Useful Field of View Test, the Symbol Digit Modalities Test, the Trail Making Test, the Controlled Oral Word Association Test, the Digit Vigilance Test, and the Stroop Color and Word Test. We have seven hypotheses for these primary outcome measures which will be tested using residualized change score multiple linear regression models for continuous outcomes, multiple logistic regression models for binary (threshold change) outcomes, and Poisson or negative binomial regression models for count measures. We will also evaluate the effects on several secondary outcomes, including HRQoL, healthcare utilization, depressive symptoms, functional abilities, perceived stress, self-efficacy, and sense of control. Finally, we will conduct stratified analysis among participants aged 50-64 years old, and separately among those aged 65 years or older in order to determine whether the effect size of the speed of processing intervention varies by age group. PUBLIC HEALH RELEVANCE: Although some degree of gradual cognitive decline is nearly universal and a normal part of the aging process, previous research by our group has shown that age-related cognitive decline is amenable to intervention. Building on speed of processing theory, we propose to extend and expand the findings from the NIH-funded, multi-site Advanced Cognitive Training for Independent and Vital Elderly (ACTIVE) study by using a newly developed, value-added version of the speed of processing software that can be used on virtually any home computer without supervision. When shown to be at least as efficacious as the original, the value-added version of the speed of processing software will then be ready for widespread implementation among adults aged 50 years old or older to reduce and/or prevent the risk of disability driven by age-related cognitive decline.